Finding count of records in given interval - sql-server

I have a table with user logins, and I want to find the user who logged in to the site more than 3 times in a 5 day period.
For example my table is:
id | user_id | login_date
---+---------+--------------
1 | 10 | 10.1.2014 00:00
2 | 10 | 11.1.2014 10:10
3 | 12 | 11.1.2014 11:00
4 | 10 | 11.1.2014 12:00
5 | 12 | 12.1.2014 00:00
6 | 10 | 13.1.2014 10:00
7 | 12 | 18.1.2014 00:00
8 | 12 | 22.1.2014 09:00
For this example table, I want to choose user_id 10 because he/she logged in more than 3 times in a 5 day period.
Could you please help me for that?
edit: i forgot to mention that database is sql server 2008

You could do it by self joining the table on user_id where the JOIN takes records that are within 5 days of the record it is joining to like so:
CREATE TABLE #login
(
id INT ,
user_id INT ,
login_date DATETIME
)
INSERT INTO #login
( id, user_id, login_date )
VALUES ( 1, 10, '2014-01-10 00:00' ),
( 2, 10, '2014-01-11 10:10' ),
( 3, 12, '2014-01-11 11:00' ),
( 4, 10, '2014-01-11 12:00' ),
( 5, 12, '2014-01-12 00:00' ),
( 6, 10, '2014-01-13 10:00' ),
( 7, 12, '2014-01-18 00:00' ),
( 8, 12, '2014-01-22 09:00' )
SELECT t1.user_id ,
t1.login_date AS FirstDateInLoginPeriod ,
COUNT(t2.user_id) AS LoginCount
FROM #login t1
INNER JOIN #login t2 ON t2.user_id = t1.user_id
AND t2.login_date
BETWEEN t1.login_date AND DATEADD(DAY, 5, t1.login_date)
GROUP BY t1.user_id ,
t1.login_date
HAVING COUNT(t2.user_id) > 3
DROP TABLE #login
Produces:
user_id FirstDateInLoginPeriod LoginCount
----------------------------------------------
10 2014-01-10 00:00:00.000 4

If you are using SQL Server 2012+ then you can use LEAD window function to calculate difference in days between any 3 consecutive records:
select distinct USER_ID
from (
select USER_ID,
datediff(d, login_date,
LEAD(login_date, 2) OVER (PARTITION BY user_id
ORDER BY login_date)) as diffDates
from users ) t
where t.diffDates <= 5
Then simply select those USER_IDs for which the difference in days is equal to or less than 5.
SQL Fiddle Demo
If you are using SQL Server 2008 or 2005, then you can use ROW_NUMBER in conjuction with a self join in order to simulate LEAD function:
;WITH CTE AS (
SELECT id, user_id, login_date,
ROW_NUMBER() OVER (PARTITION BY user_id ORDER BY login_date) AS rn
FROM users
)
SELECT DISTINCT user_id
FROM (
SELECT c1.user_id,
DATEDIFF(d, c1.login_date, c2.login_date) AS diffInDays
FROM CTE AS c1
INNER JOIN CTE AS c2 ON c1.user_id = c2.user_id AND c1.rn = c2.rn - 2
) t
WHERE t.diffInDays <= 5
diffInDays in the above query is essentially the rolling difference in days between any 3 consecutive user logins.
SQL Fiddle Demo

You can use following query to get your result.
As you can see you can use DATEADD function for calculating reference date and use having to filter logins count to be more or equal 3
Select count(*),user_id
from table_name
where login_date>=DATEADD (day,-5,GETDATE())
group by user_id
having COUNT(*) >=3

Related

T-SQL - 3 month moving sum - preceding null values

Using SQL Server 2016. I have the following data table (sample)
Target Date Total
-----------------
2018-01-24 1
2018-02-28 1
2018-03-02 1
2018-03-08 1
2018-03-15 1
2018-03-30 1
2018-04-16 1
2018-04-18 1
2018-04-30 1
I would like to get to get a 3 month moving sum (grouping is by month):
Target Date Total_Sum
-----------------------
2018-01-01 1
2018-02-01 2
2018-03-01 6
2018-04-01 8
Ok, this should get the answer you want. Firstly you need to total the value your months, then you can do a running total for the last 3 months:
CREATE TABLE SampleTable (TargetDate date, Total int);
GO
INSERT INTO SampleTable
VALUES ('20180124', 1),
('20180228', 1),
('20180302', 1),
('20180308', 1),
('20180315', 1),
('20180330', 1),
('20180416', 1),
('20180418', 1),
('20180430', 1);
GO
SELECT *
FROM SampleTable;
GO
WITH Months AS (
SELECT DATEADD(MONTH,DATEDIFF(MONTH, 0, TargetDate),0) AS TargetMonth, SUM(Total) AS MonthTotal
FROM SampleTable
GROUP BY DATEADD(MONTH,DATEDIFF(MONTH, 0, TargetDate),0))
SELECT TargetMonth,
SUM(MonthTotal) OVER (ORDER BY TargetMonth ROWS BETWEEN 2 PRECEDING AND CURRENT ROW) AS Last3Months
FROM Months;
GO
DROP TABLE SampleTable;
GO
Pls try the below code
;WITH CTE(TargetDate,Total)
AS
(
SELECT '2018-01-24', 1 UNION ALL
SELECT '2018-02-28', 1 UNION ALL
SELECT '2018-03-02', 1 UNION ALL
SELECT '2018-03-08', 1 UNION ALL
SELECT '2018-03-15', 1 UNION ALL
SELECT '2018-03-30', 1 UNION ALL
SELECT '2018-04-16', 1 UNION ALL
SELECT '2018-04-18', 1 UNION ALL
SELECT '2018-04-30', 1
)
SELECT STUFF(TargetDate,9,2,'01') AS TargetDate
,Total_Sum
FROM
(
SELECT TargetDate,Total_Sum
,ROW_NUMBER()OVER(PARTITION BY Total_Sum ORDER BY TargetDate) AS Seq
FROM
(
SELECT TargetDate
,SUM(Total )OVER(ORDER BY MONTH(TargetDate) ) AS Total_Sum
FROM CTE
)dt
)fnl
WHERE Seq=1
Result
TargetDate Total_Sum
---------------------
2018-01-01 1
2018-02-01 2
2018-03-01 6
2018-04-01 9

T-SQL create multiply records from one records

I have a cost record and I would like to create N records from it.
The children records have some different parameters.
For example:
The parents record:
date | amount | duration
20170201 | 5000 | 5 months
The children records:
date | amount | duration
20170301 | 1000 | 1 months
20170401 | 1000 | 1 months
20170501 | 1000 | 1 months
20170601 | 1000 | 1 months
20170701 | 1000 | 1 months
How can I do this without iteration? Without cursor or while?
Following SQL CTE query could be used based on Abdul's solution
/*
Create Table PARENT (PARENT_DATE DATE, PARENT_AMOUNT DECIMAL(18,2),PARENT_MONTH INT)
INSERT INTO PARENT SELECT '20170201',5000 ,5
INSERT INTO PARENT SELECT '20180601',120 ,3
*/
;WITH CTE_CHILD
AS (
SELECT
Parent_Date,
Parent_Amount,
Parent_Month,
DateAdd(Month, 1, Parent_Date) as Child_Date,
Parent_Amount/Parent_Month AS Child_Amount,
1 AS Child_Duration
FROM Parent
UNION ALL
SELECT
Parent_Date,
Parent_Amount,
Parent_Month,
DateAdd(Month, 1, Child_Date) as Child_Date,
Child_Amount,
Child_Duration
FROM CTE_CHILD
WHERE
DateAdd(Month, 1, Child_Date) <= DateAdd(Month, Parent_Month, Parent_Date)
)
SELECT
Child_Date,
Child_Amount,
Child_Duration
FROM CTE_CHILD
assuming you have a table like below:
create table tblRecords ( date int, amount money, duration int);
insert into tblRecords values
(20170201,5000,5),
(20180101,9000,3);
you can use a query like below:
select
date= date + r*100
,amount= amount/duration
,duration =1
from tblRecords
cross apply
(
select top (select duration)
r= row_number() over(order by (select null))
from
sys.objects s1
cross join
sys.objects s2
) h
see working demo
One method is CTE.
DECLARE #PARENT AS TABLE
(PARENT_DATE DATE, PARENT_AMOUNT DECIMAL(18,2),PARENT_MONTH INT)
INSERT INTO #PARENT
SELECT '20170201',5000 ,5
;WITH CTE_CHILD
AS (
SELECT DATEADD(MONTH,1,PARENT_DATE) AS CHILD_DATE
,PARENT_AMOUNT/PARENT_MONTH AS CHILD_AMOUNT
,1 AS CHILD_DURATION
FROM #PARENT
WHERE DATEADD(MONTH,1,PARENT_DATE) <= DATEADD(MONTH,PARENT_MONTH,PARENT_DATE)
UNION ALL
SELECT DATEADD(MONTH,1,CHILD_DATE)
,PARENT_AMOUNT/PARENT_MONTH
,1
FROM CTE_CHILD
INNER JOIN #PARENT ON DATEADD(MONTH,1,CHILD_DATE) <= DATEADD(MONTH,PARENT_MONTH,PARENT_DATE)
)
SELECT * FROM CTE_CHILD
option (maxrecursion 0)
Output:-
CHILD_DATE CHILD_AMOUNT CHILD_DURATION
2017-03-01 1000.0000000000000 1
2017-04-01 1000.0000000000000 1
2017-05-01 1000.0000000000000 1
2017-06-01 1000.0000000000000 1
2017-07-01 1000.0000000000000 1

SQL Select Sequential Dates with Additional Lookup Values

I am trying to grab a series of dates and the corresponding values (if any) that exist in my database.
I have two parameters - today (date using getDate()) - and a number of days (integer). For this example, I'm using the value 10 for the days.
Code to get the sequential dates for 10 days after today:
SELECT top 10 DATEADD(DAY, ROW_NUMBER()
OVER (ORDER BY object_id), REPLACE(getDate(),'-','')) as Alldays
FROM sys.all_objects
I now need to look up several values for each day in the sequential days code, which may or may not exist in the time table (we assume 8 hours for all dates, unless otherwise specified). The lookup would be on the field recordDateTime. If no "hours" value exists in the table cap_time for that date, I need to return a default value of 8 as the number of hours. Here's the base query:
SELECT u.FullName as UserName, d2.department,
recordDateTime, ISNULL(hours,8) as hours
FROM cap_time c
left join user u on c.userID = u.userid
left join dept d2 on u.deptID = d2.DeptID
WHERE c.userid = 38 AND u.deptID = 1
My end result for the next 10 days should be something like:
Date (sequential), Department, UserName, Number of Hours
I can accomplish this using TSQL and a temp table, but I'd like to see if this can be done in a single statement. Any help is appreciated.
Without any DDL or sample data it's hard to determine exactly what you need.
I think this will get you pretty close (note my comments):
-- sample data
------------------------------------------------------------------------------------------
DECLARE #table TABLE
(
fullName varchar(10),
department varchar(10),
[hours] tinyint,
somedate date
);
INSERT #table VALUES
('bob', 'sales', 5, getdate()+1),
('Sue', 'marketing', 3, getdate()+2),
('Sue', 'sales', 12, getdate()+4),
('Craig', 'sales', 4, getdate()+8),
('Joe', 'sales', 18, getdate()+9),
('Fred', 'sales', 10, getdate()+10);
--SELECT * FROM #table
;
-- solution
------------------------------------------------------------------------------------------
WITH alldays([day]) AS -- logic to get your dates for a LEFT date table
(
SELECT TOP (10)
CAST(DATEADD
(
DAY,
ROW_NUMBER() OVER (ORDER BY object_id),
getdate()
) AS date)
FROM sys.all_objects
)
SELECT d.[day], t.fullName, department, [hours] = ISNULL([hours], 8)
FROM alldays d
LEFT JOIN #table t ON d.[day] = t.somedate;
Results:
day fullName department hours
---------- ---------- ---------- -----
2017-04-12 bob sales 5
2017-04-13 Sue marketing 3
2017-04-14 NULL NULL 8
2017-04-15 Sue sales 12
2017-04-16 NULL NULL 8
2017-04-17 NULL NULL 8
2017-04-18 NULL NULL 8
2017-04-19 Craig sales 4
2017-04-20 Joe sales 18
2017-04-21 Fred sales 10
Maybe a subquery and the in statement, like:
SELECT u.FullName as UserName, d2.department,
recordDateTime, ISNULL(hours,8) as hours
FROM cap_time c
left join user u on c.userID = u.userid
left join dept d2 on u.deptID = d2.DeptID
WHERE c.userid = 38 AND u.deptID = 1 and recordDateTime in
(SELECT top 10 DATEADD(DAY, ROW_NUMBER()
OVER (ORDER BY object_id), REPLACE(getDate(),'-','')) as Alldays
FROM sys.all_objects)

Finding the Datediff between Records in same Table

IP QID ScanDate Rank
101.110.32.80 6 2016-09-28 18:33:21.000 3
101.110.32.80 6 2016-08-28 18:33:21.000 2
101.110.32.80 6 2016-05-30 00:30:33.000 1
I have a Table with certain records, grouped by Ipaddress and QID.. My requirement is to find out which record missed the sequence in the date column or other words the date difference is more than 30 days. In the above table date diff between rank 1 and rank 2 is more than 30 days.So, i should flag the rank 2 record.
You can use LAG in Sql 2012+
declare #Tbl Table (Ip VARCHAR(50), QID INT, ScanDate DATETIME,[Rank] INT)
INSERT INTO #Tbl
VALUES
('101.110.32.80', 6, '2016-09-28 18:33:21.000', 3),
('101.110.32.80', 6, '2016-08-28 18:33:21.000', 2),
('101.110.32.80', 6, '2016-05-30 00:30:33.000', 1)
;WITH Result
AS
(
SELECT
T.Ip ,
T.QID ,
T.ScanDate ,
T.[Rank],
LAG(T.[Rank]) OVER (ORDER BY T.[Rank]) PrivSRank,
LAG(T.ScanDate) OVER (ORDER BY T.[Rank]) PrivScanDate
FROM
#Tbl T
)
SELECT
R.Ip ,
R.QID ,
R.ScanDate ,
R.Rank ,
R.PrivScanDate,
IIF(DATEDIFF(DAY, R.PrivScanDate, R.ScanDate) > 30, 'This is greater than 30 day. Rank ' + CAST(R.PrivSRank AS VARCHAR(10)), '') CFlag
FROM
Result R
Result:
Ip QID ScanDate Rank CFlag
------------------------ ----------- ----------------------- ----------- --------------------------------------------
101.110.32.80 6 2016-05-30 00:30:33.000 1
101.110.32.80 6 2016-08-28 18:33:21.000 2 This is greater than 30 day. Rank 1
101.110.32.80 6 2016-09-28 18:33:21.000 3 This is greater than 30 day. Rank 2
While Window Functions could be used here, I think a self join might be more straight forward and easier to understand:
SELECT
t1.IP,
t1.QID,
t1.Rank,
t1.ScanDate as endScanDate,
t2.ScanDate as beginScanDate,
datediff(day, t2.scandate, t1.scandate) as scanDateDays
FROM
table as t1
INNER JOIN table as t2 ON
t1.ip = t2.ip
t1.rank - 1 = t2.rank --get the record from t2 and is one less in rank
WHERE datediff(day, t2.scandate, t1.scandate) > 30 --only records greater than 30 days
It's pretty self-explanatory. We are joining the table to itself and joining the ranks together where rank 2 gets joined to rank 1, rank 3 gets joined to rank 2, and so on. Then we just test for records that are greater than 30 days using the datediff function.
I would use windowed function to avoid self join which in many case will perform better.
WITH cte
AS (
SELECT
t.IP
, t.QID
, LAG(t.ScanDate) OVER (PARTITION BY t.IP ORDER BY T.ScanDate) AS beginScanDate
, t.ScanDate AS endScanDate
, DATEDIFF(DAY,
LAG(t.ScanDate) OVER (PARTITION BY t.IP ORDER BY t.ScanDate),
t.ScanDate) AS Diff
FROM
MyTable AS t
)
SELECT
*
FROM
cte c
WHERE
Diff > 30;

Find the min and max dates between multiple sets of dates

Given the following set of data, I'm trying to determine how I can select the start and end dates of the combined date ranges, when they intersect with each other.
For instance, for PartNum 115678, I would want my final result set to display the date ranges 2012/01/01 - 2012/01/19 (rows 1, 2 and 4 combined since the date ranges intersect) and 2012/02/01 - 2012/03/28 (row 3 since this ones does not intersect with the range found previously).
For PartNum 213275, I would want to select the only row for that part, 2012/12/01 - 2013/01/01.
Edit:
I'm currently playing around with the following SQL statement, but it's not giving me exactly what I need.
with DistinctRanges as (
select distinct
ha1.PartNum "PartNum",
ha1.StartDt "StartDt",
ha2.EndDt "EndDt"
from dbo.HoldsAll ha1
inner join dbo.HoldsAll ha2
on ha1.PartNum = ha2.PartNum
where
ha1.StartDt <= ha2.EndDt
and ha2.StartDt <= ha1.EndDt
)
select
PartNum,
StartDt,
EndDt
from DistinctRanges
Here are the results of the query shown in the edit:
You're better off having a persisted Calendar table, but if you don't, the CTE below will create it ad-hoc. The TOP(36000) part is enough to give you 10 years worth of dates from the pivot ('20100101') on the same line.
SQL Fiddle
MS SQL Server 2008 Schema Setup:
create table data (
partnum int,
startdt datetime,
enddt datetime,
age int
);
insert data select
12345, '20120101', '20120116', 15 union all select
12345, '20120115', '20120116', 1 union all select
12345, '20120201', '20120328', 56 union all select
12345, '20120113', '20120119', 6 union all select
88872, '20120201', '20130113', 43;
Query 1:
with Calendar(thedate) as (
select TOP(36600) dateadd(d,row_number() over (order by 1/0),'20100101')
from sys.columns a
cross join sys.columns b
cross join sys.columns c
), tmp as (
select partnum, thedate,
grouper = datediff(d, dense_rank() over (partition by partnum order by thedate), thedate)
from Calendar c
join data d on d.startdt <= c.thedate and c.thedate <= d.enddt
)
select partnum, min(thedate) startdt, max(thedate) enddt
from tmp
group by partnum, grouper
order by partnum, startdt
Results:
| PARTNUM | STARTDT | ENDDT |
------------------------------------------------------------------------------
| 12345 | January, 01 2012 00:00:00+0000 | January, 19 2012 00:00:00+0000 |
| 12345 | February, 01 2012 00:00:00+0000 | March, 28 2012 00:00:00+0000 |
| 88872 | February, 01 2012 00:00:00+0000 | January, 13 2013 00:00:00+0000 |

Resources